• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 55 subscribers
  • Views 1036 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CC2652R: Sensitivity when using Johanson balun

Michael O'Hearn
Intellectual 270 points
Part Number: CC2652R


Hi,

I'm using the CC2652 for 802.15.4 application (2.4 GHz). Our application circuit has a measured RX sensitivity around -86 to -88 dBm. TX power is in agreement with the selected settings in Smart RF studio. In comparison, we measure sensitivity around -98 dBm on TI 2652 eval board. One of the difference is that our applicaiton circuit is using the Johanson balun from App Report SWRA572 while the TI Eval board is using a lumped element balun.

One thing to note is that the TI Eval board's RSSI level is around -106 dBm when placed in Continuous RX mode and no signal applied. Our application circuit shows an RSSI level at -99 dBm when in Continuous RX mode and no signal applied.  When I remove the Johanson balun from our application circuit, RSSI decreases to -106 level. Sensitivity on our application circuit also improves with the Johanson balun removed even though we connect to one side of the CC2652's differential input.

It seems that the Johanson balun solution is causing the sensitivity problems. Could it be that it may not be optimally matched for best sensitivity performance. Our design has followed the App Report's layout recommendations. That report is from 2017, which was written before the CC2652 was available. Does that report still apply even though it is listed for CC26XX devices? Does that balun work with the CC2652 when wanting achieve the best sensitivity? Does it require match component(s) or a layout change? Is it better to use a lumped element balun?

Please let me know.

With regards

Michael 

  

  • 0
    TI__Guru**** 317180 points
    CC26x2 and CC26x0 should require the same load impedance. I can do a quick schematic and layout review to see if I see something that could explain the result (send me a friend request if you don't want to post this open on the forum)

    Is this conducted or radiated measurement? If the latter, please do a conducted measurement since the antenna could impact the result.
  • 0 in reply to TER
    Intellectual 270 points

    This is a conducted measurement.

    I need to check about sending the design files. I can say it comes out differentially from the CC2652 and to the balun. The line impedance on the balanced side is approximately 100 ohm and on the unbalanced side approximately 50 ohms.

    I've looked through the datasheet for the impedance of the RF port, but can't find anything. Also checked the Johanson Application Report. It too doesn't mention the actual impedance.

    What is the impedance in RX mode, in TX mode?

    Michael

  • 0 in reply to Michael O'Hearn
    TI__Guru**** 317180 points
    If you are not able to send me the full design, would you be able to send me a close up of the routing between the chip and the IPC?

    When you are using an IPC you don't need to know anything about the impedance but it's important to follow the reference design with regards to the length of the traces between the chip and the IPC to get the correct matching.
  • 0 in reply to TER
    Intellectual 270 points

    Hi,

    You should see the connection between the CC2652 and the balun.  The distance is 1.1 mm. The App report shows 1.0 mm. Not so much of a difference.

    I have removed the copper grounding close to the diff traces to see if it would help. It didn't.

    Also, I have measured the impedance of the CC2652. Then set in a lumped element balun based on that impedance. There was no change. To me it indicates it must be something else than the balun. We're investigating the power around the device.

    Do you have any other suggestion?

    Michael

  • 0 in reply to Michael O'Hearn
    TI__Guru**** 317180 points
    Could you include the vias and the ground plane to the plot?

    Also, I would assume you are using the internal DCDC. Could you try to switch to use the internal LDO instead since I have seen cases where poor layout around the DCDC components reduce the sensitivity and testing with the LDO is effective way to test if this is the case.
  • 0 in reply to TER
    TI__Guru**** 317180 points
    Any progress on this?